¹ RE i (1894), 1846. The letter in question is IX, no, p. 265 in Seeck's edition (MGH AA VI, 1, 1883); I shall refer to Seeck's preface to his edition simply as praef.

² Thompson, E. A., The Historical Work of Ammianus Marcellinus (1947), 18Google Scholar (referred to hereafter simply by author's name and page); Pighi, G. B., Ammiani Marcellini Rerum Gestarum capita selecta (1948), xiGoogle Scholar; Pack, R. A., TAPhA LXXXIV (1953), 187, n. 22Google Scholar; Naudé, C. P. T., Ammianus Marcellinus in die lig van die antieke geskiedskrywing (1956), 35f.Google Scholar

³ Wytzes, J., Der Streit um den Altar der Victoria (1936), 34Google Scholar; Thompson, E. A., Greece and Rome XXXII(1942), 134Google Scholar; de Jonge, P., Ut Pictura Poesis: studialatina Petro Iohanni Enk septuagenario oblata (1955), 101Google Scholar; Rolfe, J. C., Loeb edition of Ammianus, I (1956), xivGoogle Scholar; Chastagnol, A., [La] Préfecture [urbaine à Rome sous le bas-empire (1960)], 12 f.Google Scholar; Tränkle, H., Antikeund Abendland XI (1962), 26 f.Google Scholar; and most recently Momigliano, A. D. in The Conflict between Paganism and Christianity in the Fourth Century (1963), 97.Google Scholar

⁴ [A] Conflict [of Ideas in the Late Roman Empire, 1952], 4, 41; Alföldi's view of Ammianus has not generally been accepted: cf., for example, Baynes, N. H., JRS XLIII (1953), 169.Google Scholar

⁵ Zur Geschichtsschreibung und Weltanschauung des Ammianus Marcellinus, Klio, Beiheft xvi (N.F. 111), 1923 (referred to hereafter simply by author's name and page), 7 f., following a suggestion of Peter, H., Die Geschichtliche Literatur über die römische Kaiser-zeit (1897) 11, 33–4Google Scholar, that Symmachus' correspondence was edited for Christian readers, and letters written to the usurpers Maximus and Eugenius and their supporters were omitted or published without the names of their addressees. It is significant that Peter did not see fit to repeat this suggestion in his much more detailed study of the circumstances of the publication of Symmachus', letters, ‘Die Brief in der römischen Literatur,’ Abh. kön. sächs. Akad. d, Wiss. xx (1901), 135–49Google Scholar, as pointed out by McGeachy, J. A. Jr, CPh XLIV (1949), 222Google Scholar, who shows by a careful analysis that Peter's earlier view is quite untenable.

⁶ Römische Kinderkaiser (1951), 73, n. 2: also Jannacone, S., Ammiano Marcellino, profilo storicocritico (1960), 26 f.Google Scholar, and, implicitly, Straub, J., Studien zur Historia Augusta, = Diss. Bernenses I, 4 (1952), 141.Google Scholar

⁷ HThR XLVII (1954), 320.

⁸ p. 108. I should not like my criticisms of this one aspect of Thompson's book to imply that I do not value as a whole his admirable and important study of Ammianus.

⁹ p. 110.

¹⁰ RE I, 1846, and cf. also Alföldi, Conflict, 108 f.

¹¹ For numerous other examples from the subscriptions dating ultimately from ‘editions’ of the fourth to sixth centuries see the classic study of Jahn, O., Ber. sächs. Gesell. Wiss. III (1851), 327 f.Google Scholar But as all extant manuscripts of Ammianus are descended, directly or indirectly, from one very defective and lacunose archetype, as recently confirmed in a detailed study by W. Seyfarth as a Vorarbeit to a projected new edition of Ammianus, (‘Der codex Fuldensisu. der Codex E des Ammianus Marcellinus’, Abh. d.deutsch. Akad. d. Wiss., Klassef. Sprachen Literatur u. Kunst, 1962, 11)Google Scholar, it is possible, as Ensslin suggested (p. 8), that the initials were missing in this archetype.

¹² ‘itane … paginis tuis lenocinia aevi praesentisanteferas’ (Ep. IV, 30).

¹³ Ensslin, p. 25 f.

¹⁴ Seeck, praef. cii, with stemma at xci; see also now the revised stemma of Chastagnol, A., [Les] Fastes [de la Préfecture de Rome au bas-empire, 1962] 294.Google Scholar On Ammianus and the Anicii, especially Probus, cf. Alföldi, Conflict, 23 f.

¹⁵ Pighi accepts the invadas of the MSS, but we should surely follow Seeck in adopting Juretus' easy emendation invideas in order to give ‘saeculo nostro’ a construction.

¹⁶ o.c. (n. 2), xi.

¹⁷ o.c. (n. 2), 109.

¹⁸ Not accepted by Baehrens, Bursian's Jahresb. CClll (1925), 83, or by Stein, , Bas-Empire I (1959), 534, n. 146.Google Scholar

¹⁹ Ep. I, 14, 2; I, 15, 2 (both to Ausonius); 111, 44, 1 (to Siburius); 111, 47 (to Eutropius); VIII, 23, 1 (to Marcianus).

²⁰ I use the word ‘Latin’ merely as a convenient brachylogy for a Latin-speaking citizen of the Empire.

²¹ Kroll, W., ‘De Q. Aurelii Symmachi studiis Graecis et Latinis,’ Breslauer Phil. Abh. VI, 2 (1891), 6–16Google Scholar, has shown just how small Symmachus' acquaintance with Greek was; he even doubts whether he had actually read any Homer. Cf. also Courcelle, P., Les Lettres grecques en Occident de Macrobe à Cassio-dore ² (1948), 4f.Google Scholar

²² I am inclined to agree with Kroll, o.c. (n. 21), 12, n. 1, that Seeck was wrong to exclude ‘id est aequi-manum’ as a gloss. Kroll compares Jerome, In Daniel 12, 5 for a similar ‘gloss’ by the author himself. One might add that the word aequimanus occurs nowhere in Latin literature before Symmachus except in Ausonius. This would not be the first time Symmachus had decked out his correspondence with flosculi borrowed from his Gallic friend; cf. ‘delenifica et suada facundia’, a phrase which Ausonius had used in a letter to Symmachus (Ep. XVII), who promptly used it in one of his own letters (Ep. III, 6, 1). (Aequimanus is, however, once found as a gloss; cf. Corp. Gloss. I, 402, 29.)

²³ See Momigliano, A. D., Secondo Contribute alla storia degli studi classici (1960), 145 f.Google Scholar

²⁴ On which see most recently Straub, J., Heidnische Geschichtsapologetik in der Christlichen Spätantike: Untersuchungen über Zeit und Tendenz der Historia Augusta (Antiquitas Reihe IV, Bd.1, Bonn, 1963)Google Scholar, and my article ‘Literary allusions in the Historia Augusta’ in Hermes XCII (1964).

²⁵ HThR XLVII (1954), 320.

²⁶ o.c. (n. 2), XI.

²⁷ He records an examination of his letters after the conspiracy in 371 (Or. I, 175 f.) and there is a complete lacuna in his correspondence between 365 and 388, since he was suspected of supporting the usurpers Procopius and Maximus (cf. Norman, A. F., JHS LXXX, 1960, 125Google Scholar). His grandfather and greatuncle had both been executed for alleged treasonable activities (Sievers, , Das Leben des Libanius, 1868, 5Google Scholar).

²⁸ Seeck, praef. XXVI; Peter, ‘Die Brief …’, 148; McGeachy, o.c. (n. 5), 222 f.

²⁹ But probably Protadius, as there are reasons for supposing that all the letters written to him and his two brothers were preserved together: Seeck, praef. XXIV, cxliii.

³⁰ Seeck, praef. xxvi.

³¹ Seeck, praef. cxxvi. This is the same Naucellius some of whose poetic works were discovered only a dozen years ago in the Vatican: cf. F. Munari, Epigrammata Bobiensia (1955); Speyer, W., Naucellius u. seine Kreis, = Zetemata, Heft 21 (1959)Google Scholar; and Speyer's recent Teubner edition (1963).

³² Ep. 1063, ed. Förster XI 186: writing such commendatory letters was, in fact, one of Libanius' specialities (cf. Liebeschuetz, W., Historia XII, 1963, 228 f.Google Scholar).

³³ Assuming that the Marcellinus addressed in this letter is the historian: there are several Ammiani and Marcellini among Libanius' correspondents, cf. Seeck, , [Die] Briefe [des Libanius zeitlich geordnet, = TuU xxx (1906)], 57 f.Google Scholar, 201 f. I do not know on what grounds Thompson, elsewhere (CQ XXXVIII, 1944, 52)Google Scholar cites this letter as evidence that Ammianus' history was already being read at Antioch at this date.

³⁴ o.c. (n. 2), vii. It may be relevant to note that St. John Chrysostom, who was certainly a pupil of Libanius, shared his master's dislike of the Byzantine accentual clausula (Maas, Greek Metre, tr. Lloyd Jones, 1962, § 23), whereas Ammianus observed it most strictly (see the preface to Clark's edition, 1910, vi f.).

³⁵ Sievers, Leben des Libanius, 61 f.

³⁶ Bidez, , Vie de l'empereur Julien (1930), 55 f.Google Scholar, 277 f.; in medieval legend Libanius was Julian's seneschal (ib. 339).

³⁷ Pack, , TAPhA LXXXII (1951), 183.Google Scholar

³⁸ Thompson, 2, 81, 83, 85, 128 f.

³⁹ On Julian's measures see Jonge, De, Mnemosyne I ⁴ (1948), 238 f.CrossRefGoogle Scholar

⁴⁰ Nuovi studi ammianei (1936), 30, 147 f.

⁴¹ CIL VI 1698; cf. also Chastagnol, Fastes, 159 f.

⁴² p. 80, following Seeck, , Hermes XLI (1906), 533, n. 1Google Scholar; Seeck discusses Maximus' career in detail (making the same error) at Briefe 206, Maximus 11.

⁴³ Beiträge zur alten Gesch. (later Klio) II (1902), 245, no. 32. Seeck's error is repeated by Ensslin, RE Suppl. V, 671, s.v. Maximus 63 (though not at RE XIV. 2, 2543). The only correct account of the career of Maximus is to be found in Chastagnol, Fastes, 154–6.

⁴⁴ Paulinus of Nola, Ep. 29, 6.

⁴⁵ Murphy, F. X., Traditio v (1947), 59 f.CrossRefGoogle Scholar

⁴⁶ See the stemma gentis maternae Galli Caesaris ap. Chastagnol, Fastes, 296.

⁴⁷ Bidez, Oeuvres de Julien I. 2, Lettres (1932), 31: cf. especially Eutropius, , Brev. X, 15–16.Google Scholar

⁴⁸ Seeck, Briefe, 190 f: see also n. 62. Ammianus thought it worthy of especial note that Valentinian did not reward his relatives with high office (XXX, 9, 2)—except, of course, for his brother.

⁴⁹ Seeck, Briefe 252.

⁵⁰ Peter the patrician, fr. 16, FHG IV 190.

⁵¹ Pliny, NH XXXVI, 181; see Rougé, J., REA lxiii (1961), 63 f.Google Scholar

⁵² Chastagnol, Préfecture, 348.

⁵³ Senatorial landowners naturally found it irksome to subsidize the wine supply of the capital: cf. Mazzarino, S., Aspetti sociali del quarto secolo (1951), 381Google Scholar, n. 6.

⁵⁴ Rougé, o.c. (n. 51), 72 f.

⁵⁵ Cod. Theod. XI 2, 2.

⁵⁶ Symmachus Ep. 11, 38, if Seeck is correct (praef. xliii, adn. 96, and cxxiii) in referring this letter to the incident; Rougé takes over Seeck's conclusions without indicating that they were conjectural.

⁵⁷ o.c. (n. 54), 381, n. 6.

⁵⁸ Romano, D., Simmaco (1955), 28, n. 45.Google Scholar

⁵⁹ Septimius Acindynus, probably a senator by birth (Symmachus, Ep. I, 1, 2), was vicarius in Spain some time between 317–326 (CIL II 4107). Lucerius Verinus, vicarius in Africa in 318–21 (Cod. Theod. IX, 15, 1) and city prefect in 323–4 (Chron. Min. I, 67), was probably a senator by birth, as only such seem to have been appointed prefect by Constantine: Chastagnol, Préfecture, 403, supposing him an eques on the strength of his vicariate in Africa, concludes that his prefecture was a ‘choix exceptionel d'un nonnoble’. Aconius Catullinus, vicarius in Africa in 338–9 (Cod. Theod, XII, 1, 24: the year after Con-stantine's death) was also very probably a noble. Stein's theory, (Bas-Empire I, 70) that it was deliberate policy to make senatorial governors dependent on vicarii of lower rank has not generally been accepted (cf. Seston, W., Dioclétien et la Tétrarchie I, 1946, 338Google Scholar), and would seem to be quite baseless if the above examples are considered valid. I owe the references, and the point, to Professor Jones, A. H. M.. (See now his Later Roman Empire (1964) 1, 107Google Scholar; 11, 527.)

⁶⁰ Son according to Seeck praef. xli, grandson according to Chastagnol, Fastes, 112, with a revised stemma of the Symmachi at p. 294. On the names of this Symmachus see Chastagnol, ib. 112.

⁶¹ Cf. Alföldi, , The Conversion of Constantine and Pagan Rome (1948), 119, n. 2.Google Scholar The civil character ofthe vicariate is emphasized by Ensslin, , CAH XII (1939). 394.Google Scholar

⁶² In fact the senate met his overtures with hostility, as can be clearly seen from Ammianus XXI, 10, 7. This may have been another reason why Julian ‘passed over’ Avianius Symmachus and appointed Maximus city prefect: Symmachus, renowned as a diplomat, was probably in charge of the embassy that met Julian, and may well have indicated to Julian the senate's disapproval of his action. Under these circumstances Julian would naturally have preferred to rely on a relative than a man who probably, at this stage at any rate of the rebellion, shared the loyalty of the rest of the senate to Constantius (Ammianus l.c.). Furthermore a sarcastic reference to Symmachus in a letter written a year later (Ep. 82, p. 137 Bidez) suggests that Julian positively distrusted him.

⁶³ Studien zu Ammianus Marcellinus (Klio, Beiheft XIII, 1914), 38.

⁶⁴ On Praetextatus' religious ideas and priesthoods cf. Bloch, H., HThR XXXVIII (1945), 203 f. and 242.Google Scholar

⁶⁵ Ensslin, p. 30.

⁶⁶ Testimonia on Praetextatus, Seeck, praef. lxxxiii; Chastagnol, Fastes, 171–8.

⁶⁷ Not even Thompson's demonstration of Ammianus' bias for Ursicinus and against Gallus and Maximinus alters the fact (as Thompson himself emphasizes, p. 125) that he was in general, as Gibbon observed, ‘without the prejudices and passions which usually affect the mind of a contemporary.’

⁶⁸ All we know for certain about Ammianus' movements at this time is that he was still in Antioch in 378: Thompson p. 14.

⁶⁹ Though apparently accepted by Bidez, Vie (n. 36), 131. Why should Eusebia, Julian's protector (Amm. XV, 2, 8; 8, 3) and arranger of the match herself, have wanted to spite his wife in this manner ? See also Pack, , AJPh LXIX (1948), 307, n. 16.Google Scholar

⁷⁰ For the gaps see next note.

⁷¹ Thompson, p. 139, suspects that Ammianus deliberately omitted the names to avoid touching on awkward facts. Chastagnol, Préfecture, 428, agrees with Thompson, but Kohns, H. P., Versorgungskrisen [und Hungerrevolten im spätantiken Rom: Antiquitas Reihe I. 6 (1961)], 142Google Scholar, follows Seeck in supposing a lacuna. But then we should have to postulate another lacuna to account for the prefecture of Tanaucius Isphalangius in 374–5, likewise omitted by Ammianus (Chastagnol, Fastes, 194). There is no evidence for the prefecture of Pasiphilus ‘in the summer of 355’ mentioned by Thompson, p. 27.

⁷² Ensslin p. 28; Hartke, Kinderkaiser, 333; according to Alföldi, Conflict 128, ‘it cannot reasonably be doubted,’ but see Thompson, , CR LXVIII (1954), 63 f.Google Scholar, and McGeachy, , CPh L (1955), 281 f.Google Scholar If Alföldi is right about this, then the bias against Valentinian with which he reproaches Ammianus will reflect the bias of Flavianus rather than Ammianus' own ‘malice’ (P. 42).

⁷³ The ideals, even the pen, of Flavianus have been detected by Émilienne Demougeot in the notoriously senatorial Historia Augusta (Ant. Class. XXII, 1953, 361 f.).

⁷⁴ Warmington, B. H., ‘The career of Romanus comes Africae,’ Byz. Zeit. XLIX (1956), 55–6.Google Scholar

⁷⁵ p. 89 f.

⁷⁶ Epit. de Caes. XLVIII, 11.

⁷⁷ Cf. Maenchen-Helfen, O. J., AJPh. LXXVI (1955), 384 f.Google Scholar

⁷⁸ Prudentius, for example, throughout his detailed refutation of Symmachus' famous relatio, treats Symmachus himself with the very highest respect (Contra Symm. I 632 f.; II praef. 56 f.; II 19, 644, 760 f.).

⁷⁹ Romano, D., Simmaco (1955), 26Google Scholar, with a collection of Symmachus' references to it.

⁸⁰ ap. Laqueur, , Koch, , Weber, , Probleme d. Spätantike (1930). 35 f.Google Scholar

⁸¹ For a criticism of Laqueur's method and a different explanation of the passage see Baynes, N. H., JRS XXV (1935), 87.Google Scholar

⁸² Altgermanien (1934), 61.

⁸³ On the Orfitus affair see Chastagnol, , Annales écon. soc. civil. V (1950), 161Google Scholar; Préfecture, 343 f.; and Sinnigen, W. G., The Officium of the Urban Prefecture (1957), 49 f.Google Scholar

^83a cf. McGeachy, J. A. Jr, Q. Aurelius Symmachns and the Senatorial Aristocracy of the West, (Diss. Chicago, 1942), 188.Google Scholar

⁸⁴ Maenchen-Helfen, , AJPh. LXXVI (1955), 386 f.Google Scholar

⁸⁵ Romano, D., Claudiano (1958), 143 f.Google Scholar

⁸⁶ cf. Libanius Ep. 1063 (to Ammianus): ἐν ἐπιδείξεσι ταῖς μὲν γέγονας, ταῖς δὲ ἐσῇ τῆς συγγραφῆς εἰς πολλὰ τετμημένης καὶ τοῦ φανέντος ἐπαινεθέντος μέρος ἕτερον εἰσκαλοῦντος.

⁸⁷ Pack, , TAPhA LXXXIV (1953), 182.Google Scholar

⁸⁸ Roman Society in the Last Century of the Western Empire ² (1899), 123.

⁸⁹ Les lettres grecques (n. 21), 6, n. 2, 18.

⁹⁰ Plotin et l'Occident (1934), 196 f.

⁹¹ de Labriolle, P., La réaction paienne (1934), 189, 454 f.Google Scholar

⁹² Loyen, A., Sidoine Apollinaire et l'esprit précieux en Gaule aux derniers jours de l'empire (1943), 27.Google Scholar

⁹³ See the discussion of J. Bayet, in the preface to his Budé edition of Livy, I (1940), xcii f.

⁹⁴ kroll, o.c. (n. 21), 7f., 17 f.

⁹⁵ On whom see especially Hagendahl, H., Latin Fathers and the Classics, (1958), 284 f., 341 f., 347 f.Google Scholar On Cicero and Virgil in Ammianus see Fletcher, G. B. A., Rev de Phil, LXIII (1937), 377 f.Google Scholar, and earlier works there cited.

⁹⁶ Highet, G., Juvenal the Satirist (1954), 186Google Scholar; Griffith, J. G., Hermes XCI (1963), 107Google Scholar; and my own article in Hermes XCII (1964).

⁹⁷ Romano, Simmaco, 94–9, devotes a whole section to the topic of amicitia in Symmachus.

⁹⁸ Seeck, , RE I 1846Google Scholar; Thompson p. 14; Chastagnol, Préfecture, 268, n. 5.

⁹⁹ TAPhA LXXXIV (1953), 188–9.

¹⁰⁰ REA XXXIII (1931), 350 f.

¹⁰¹ Chastagnol, , Rev. Beige de Phil. XL (1962), 917Google Scholar, referring to an unpublished paper of P. Petit dating the speech to 356.

¹⁰² See Kohns, Versorgungskrisen (n. 71), 175.

¹⁰³ JEA XL (1954), 76 f.

¹⁰⁴ Kohns, Versorgungskrisen, 172–80; Chastagnol, Fastes, 223; Préfecture, 268; and Ruggini, Lellia, Economia e società nell' Italia annonaria (1961), 161 f.Google Scholar

¹⁰⁵ De off. min. III, 49; cf. Kohns, , Versorgungskrisen 75, 170–1.Google Scholar The prefect was directly responsible for such expulsions; see Chastagnol, Préfecture, 268 f.

¹⁰⁶ Perhaps, as Italo Lana supposes (Letteratura latina, 1963, 444), after, and as a result of, the recitation of his history.

¹⁰⁷ That there was a certain pressure to include information supplied by powerful personages—evidently resented by Ammianus—appears from XXVI, 1, 1.

¹⁰⁸ Alföldi, for example, seems to regard the words ‘Ammianus’ and ‘senatorial circles’ as interchangeable: on pp. 16–17 of bis Conflict he writes ‘senatorial circles’ twice and ‘senatorial tradition’ once, on each occasion referring simply to a passage of Ammianus.

I am grateful to Professors A. H. M. Jones and A. D. Momigliano for reading an earlier draft of this article and to Professor A. Chastagnol, who kindly communicated to me some entries from his Fastes before the book became available in this country.